Locality-Sensitive Hashing-Based Efficient Point Transformer for Charged Particle Reconstruction

TL;DR

The paper targets the heavy computational burden of charged-particle tracking under HL-LHC conditions by comparing a locality-sensitive, near-linear HEPT-based approach to a strong GNN baseline and introducing HEPTv2, an end-to-end decoder that eliminates clustering. HEPTv2 preserves the hardware-friendly, regular computations of HEPT while enabling direct track assignments, achieving ~28 ms per event on TrackML with an A100. Across pixel tracking tasks, it delivers competitive efficiency with significantly reduced latency and without post-hoc clustering. This work demonstrates a practical, scalable alternative to GNN-based pipelines for fast, real-time tracking in high-rate collider experiments.

Abstract

Charged particle track reconstruction is a foundational task in collider experiments and the main computational bottleneck in particle reconstruction. Graph neural networks (GNNs) have shown strong performance for this problem, but costly graph construction, irregular computations, and random memory access patterns substantially limit their throughput. The recently proposed Hashing-based Efficient Point Transformer (HEPT) offers a theoretically guaranteed near-linear complexity for large point cloud processing via locality-sensitive hashing (LSH) in attention computations; however, its evaluations have largely focused on embedding quality, and the object condensation pipeline on which HEPT relies requires a post-hoc clustering step (e.g., DBScan) that can dominate runtime. In this work, we make two contributions. First, we present a unified, fair evaluation of physics tracking performance for HEPT and a representative GNN-based pipeline under the same dataset and metrics. Second, we introduce HEPTv2 by extending HEPT with a lightweight decoder that eliminates the clustering stage and directly predicts track assignments. This modification preserves HEPT's regular, hardware-friendly computations while enabling ultra-fast end-to-end inference. On the TrackML dataset, optimized HEPTv2 achieves approximately 28 ms per event on an A100 while maintaining competitive tracking efficiency. These results position HEPTv2 as a practical, scalable alternative to GNN-based pipelines for fast tracking.

Figures (1)

• Figure 1: Tracking efficiency as a function of (a) transverse momentum, $p_{\text{T}}$, and pseudorapidity, $\eta$.